On April 26, 2007, a patient from Alberta, Canada, died after 9 weeks in an intensive care unit (ICU) from encephalitis caused by a rabies virus variant associated with silver-haired bats. This report summarizes the clinical course of disease in that patient, who was treated using the Milwaukee Protocol, an experimental treatment protocol similar to one used for the rabies survivor described in 2005.1 This report also describes the subsequent epidemiologic investigations by three regional public health departments in Alberta. Rabies continues to be a cause of human death in the developed and developing world. The findings in this report underscore the need for continued public education that promotes rabies prevention and postexposure prophylaxis while emphasizing the importance of bat exposure in rabies transmission.
During August 2006, a man aged 73 years was bitten by a bat on his left shoulder while sleeping at home in rural Alberta. He killed and disposed of the bat and did not seek medical attention. The patient had no history of previous rabies vaccination and became ill on February 14, 2007, when he had onset of left shoulder pain. The pain was radicular, severe, and progressive and evolved to include left hand weakness during the next few days. The man sought care at a local emergency department on February 15, 17, and 19, and was administered analgesics.
On February 21 (the seventh day of clinical illness), the patient was admitted to the local hospital with general weakness, anorexia, and dysphagia. His family described the patient as irritable and not himself. Forty-eight hours after admission, the patient had left arm myoclonus and gasping respirations, suggestive of inspiratory spasms. His illness progressed with high fever, hypoxia, hypersalivation, and a decreased level of consciousness. He required intubation and was transferred to a tertiary-care hospital ICU on February 23 (the ninth day of clinical illness) with a presumptive diagnosis of aspiration pneumonia and sepsis. The history of a previous bat bite was not obtained at that time.
A computerized tomography scan of the head on admission to the tertiary-care hospital was unremarkable. A lumbar puncture was performed, and analysis of cerebrospinal fluid (CSF) indicated no white blood cells, normal glucose, and marginally elevated protein. A chest radiograph revealed a right lower lobe infiltrate, and treatment for presumed pneumonia with broad-spectrum antibiotics was initiated. The patient continued to deteriorate with cardiac dysrhythmias, profound hemodynamic lability, opisthotonic posturing, hypersalivation, and diffuse spasticity. Because of this evolution of the patient's symptoms, rabies was considered as a possible diagnosis on February 26 (the 12th day of clinical illness). When asked about bites or other exposures, the patient's family recalled that the patient had been bitten by a bat approximately 6 months before.
A nuchal biopsy specimen and saliva sample were sent to the Canadian Food Inspection Agency in Ottawa, Ontario, where the rabies diagnosis was confirmed on March 1 (the 15th day of clinical illness). Presence of viral antigen and viral RNA was detected by direct fluorescent antibody test (DFA) and reverse transcription polymerase chain reaction (RT-PCR), respectively. Subsequently, the rabies virus RNA was typed as a variant associated with silver-haired bats (Lasionycteris noctivagans).
Rabies immune globulin was administered (1,200 units intramuscularly) on March 1. After discussion with the family regarding the diagnosis, the poor prognosis, and possible management strategies, a decision was made to initiate the Milwaukee Protocol, a recently described experimental therapy for rabies.1 This regimen involves (1) induction of therapeutic coma, (2) waiting for an adaptive immune response to evolve and neutralize and clear virus from the central nervous system and periphery, and (3) supportive antiviral and metabolic therapies. In 2004, this protocol resulted in survival and good neurologic outcome for an unvaccinated female patient aged 14 years in Milwaukee, Wisconsin.1 On March 2 (the 16th day of clinical illness), the treating physicians initiated the Milwaukee Protocol, including parenteral ketamine infusion (2 mg/kg), midazolam infusion (0-20 mg/hour), ribavirin (560 μg every 8 hours), and amantadine (200 mg once daily); the protocol was modified to include L-arginine (35 g every 24 hours), enteral administration of tetrahydrobiopterin (150 mg every 8 hours), and vitamin C (500 mg once daily) to supplement possible deficiencies and to improve cerebral blood flow autoregulation. The immunologic response and peripheral viral clearance were monitored via detection of viral RNA in saliva by quantitative RT-PCR and titration of rabies virus neutralizing antibodies in sera and CSF using a rapid fluorescent focus inhibition test.
The patient's severe hemodynamic lability improved gradually on ventilatory and low-dose pressor support. Rabies immunoglobulin G (IgG) and immunoglobulin M (IgM) were detected in serum on March 6 and in CSF on March 11, a total of 20 and 25 days, respectively, after onset of neurologic symptoms. Baseline serum and CSF tested negative for the presence of IgM and IgG against rabies virus, and subsequent development of an IgM response was thought to represent an immune response to the infection. The patient was weaned from sedation and, on April 1 (the 46th day of clinical illness), sedation was removed completely. However, no neurologic recovery occurred despite detection of low titers of virus-neutralizing antibodies (0.46-1.16 IU/mL) in CSF and normal cerebral perfusion.
Levels of virus-neutralizing antibodies in serum increased slowly and reached 0.9 IU/mL on April 24 (the 69th day of clinical illness). During the disease course, detectable rabies virus decreased markedly in the peripheral tissues, with a negative DFA on the skin biopsy and a small amount of viral RNA detected by PCR in saliva. During the same period, the patient had cardiac arrhythmias, autonomic instability, syndrome of inappropriate antidiuretic hormone secretion, hemolysis attributed to ribavirin, and ventilator-associated pneumonia.
A nuclear medicine brain death scintigraphy study revealed preserved brain perfusion; however, on April 23 (the 68th day of clinical illness), repeated magnetic resonance imaging demonstrated diffuse severe signal abnormality of the cortex, white matter, basal ganglia, and thalami. Clinical examination, including apnea testing, was consistent with brain death. After discussion with the family, life-support was withdrawn on April 26, approximately 8 weeks after initiating therapy, and the patient died. DFA staining of the autopsied brain stem and cerebral cortex demonstrated an abundance of rabies viral inclusions. These results were confirmed by RT-PCR. Microscopic examination revealed extensive and virtually complete loss of cortical neurons, whereas the cerebellum and brainstem had preservation of neurons.
In conjunction with the admitting tertiary-care hospital, the public health departments of three Alberta health regions traced the household and health-care–associated contacts of the patient starting from 1 week before onset of neurologic symptoms, a practice consistent with previous similar investigations.2 Postexposure prophylaxis (PEP) was recommended for health-care workers and close contacts of the patient with a possible exposure (defined as a bite, scratch, or exposure of nonintact skin or mucous membrane surface to saliva, CSF, tears, or brain tissue). A total of 19 contacts received PEP. All family members (the patient's wife and his two sons) were administered PEP with rabies immune globulin and vaccine. Sixteen health-care workers, who had reported exposures of mucous membranes or nonintact skin to the patient's saliva, were administered PEP; 15 (six from the primary referring hospital and nine from the tertiary-care hospital) received rabies immune globulin and vaccine. One health-care worker, who had been vaccinated previously, received 2 booster vaccine doses. To date, none of the persons who received PEP have demonstrated illness consistent with rabies.
J Johnstone, MD, L Saxinger, MD, Infectious Diseases, R McDermid, MD, S Bagshaw, MD, Critical Care, L Resch, MD, Pathology, Univ of Alberta; B Lee, MD, Alberta Provincial Public Health Laboratory; M Johnson, MD, Public Health Div, AM Joffe, MD, Occupational Health, Safety, and Wellness, Capital Health Region, Edmonton; G Benade, MD, Public Health Div, East Central Health Region, Camrose; D Johnson, MD, Public Health Div, Aspen Health Region, Westlock, Alberta; S Nadin-Davis, PhD, Canadian Food Inspection Agency, Ottawa; E Cheung, Public Health Laboratories Br, Ministry of Health and Long-Term Care, Etobicoke, Ontario, Canada. R Willoughby Jr, MD, Medical College of Wisconsin, Milwaukee, Wisconsin. R Franka, DVM, PhD, Div of Viral and Rickettsial Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, CDC.
In Canada, 24 documented human rabies cases, including the one described in this report, have occurred since 1924.2,3 Since 1970, six of the seven cases have been attributable to rabies virus variants associated with bats.2,3 Bats are an increasingly common source of human rabies in the United States, accounting for 37 (92.5%) of the 40 indigenous cases of rabies since 1990.4 Passive surveillance of bats in western Canada during 1985-1989 indicated that 4.8% of bats submitted for testing were positive for the presence of rabies virus; the prevalence has remained stable since 1965.5 The rabies virus variant associated with L. noctivagans bats in North America has been implicated in multiple indigenously acquired human rabies cases in the United States in recent years and also was responsible for a case of human rabies in Quebec, Canada, in 2000.6
After an exposure, human rabies is preventable by local wound care and administration of PEP.3,7,8 Patients with no previous rabies vaccination require rabies immune globulin and a 5-dose series of rabies vaccine.7,8 However, as the case in this report illustrates, persons are not always aware of the importance of seeking attention and PEP after bat exposures. In addition, clinicians need to recognize that a majority of patients with human rabies transmitted by bats might have no recollection of a bat bite. Thus, PEP should be considered in circumstances in which the likelihood of a bite cannot be reasonably excluded.7,8 PEP can be administered any time after an exposure, up to the onset of neurologic illness, but effectiveness of prophylaxis decreases with time; therefore, early administration of PEP is critical. After infection, the usual incubation period for rabies is 20 to 60 days, although it can vary from several days to years.8
Only one unvaccinated rabid patient (the girl in the Milwaukee case) has survived. Several other attempts to use the Milwaukee Protocol have been unsuccessful.9 Compared with the Milwaukee patient, the patient in this report (1) had advanced age; (2) had encephalitic disease with high levels of viral load in saliva and no detectable antibody response at the time of diagnosis; and (3) had received rabies immune globulin. Immune globulin administration during clinical rabies has not been demonstrated to be useful and is not part of the Milwaukee Protocol because of concerns that it might alter the kinetics of the immune response.10
Sixteen health-care workers received PEP after the public health investigation. The indication for PEP includes exposure of nonintact skin or mucous membranes to potentially infectious body fluids (e.g., saliva) or neuronal tissue; standard infection-control precautions can minimize health-care workers' risk for exposure to rabies virus.7,8 To date, no cases of transmission of rabies to persons exposed through health-care activities have been documented.
This report underscores the need for increasing public awareness of the risk for rabies after contact with bats. Underestimation of the importance of such exposures can lead to a fatal outcome. Persons bitten by a bat should immediately (1) wash the wound thoroughly with soap and water; (2) capture the animal, if this can be done safely (otherwise call local animal-control services for assistance), and submit the bat for testing; (3) report the incident to local or regional/state public health officials; and (4) visit a physician for treatment and evaluation regarding the need for PEP. Timely submission of the bat (or other possibly rabid animal) to public health officials facilitates testing for the presence of rabies virus, helps to ensure rapid administration of PEP when indicated, and minimizes the unnecessary use of PEP if the animal is not rabid.
An experimental approach to treat rabies in humans requires early diagnosis. Therefore, rabies should be included in the differential diagnosis of any unexplained acute, rapidly progressive viral encephalitis.
Rabies is a fatal but easily preventable disease that has no established effective therapy after onset of clinical disease. In addition to animal vaccination, continued public education regarding rabies exposure and timely and appropriate prophylaxis is a primary strategy for human rabies prevention.
This report is based, in part, on contributions by staff members in the Capital and Aspen Health Regions, B Aitken, East Central Health Region, Alberta; F Muldoon, M Sheen, C Fehlner-Gardiner, A Wandeler, Canadian Food Inspection Agency, Ottawa, M Shaw, T Okura, R Kandiah, Public Health Laboratories Br, Ministry of Health and Long-Term Care, Etobicoke, Ontario; and CE Rupprecht, VMD, PhD, LA Orciari, MS, M Niezgoda, MS, A Velasco-Villa, PhD, PA Yager, I Kuzmin, MD, Div of Viral and Rickettsial Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, CDC.
Human Rabies—Alberta, Canada, 2007. JAMA. 2008;299(23):2740–2742. doi:10.1001/jama.299.23.2740